Here is the result I more than half expected.
It's my fault.

This is the armature of a 24V rated single reduction Chinese gear motor for electric assist bicycles.

It's a product of Unite Motor.
Currie USA uses this model for their entry-level ebikes.

It's a great little motor; very torquey on 24V.
But, wanting more speed than 17mph, I do and did as e-bike tinkerers like to do: Added another battery for -more speed-.

Oh, wowsirs: the bike ran strong at 24mph.
For all of seven miles at full throttle
*done as a test, expecting either it'd live or die*

it died

---

The gearmotor is an under-eighty dollar unit.

--For the next trial (I must keep the 36V power and speed) I would like to force-air cool the motor.

Do you have ideas? May I post some more pictures? The motor is totally enclosed.
However, it would be easy to shoot air into and through it, by nipple connections and a rubber tube.

Evidence clues me that the motor died simply because it was run too hard.

The brushes and commutator held up beautifully.
No brush dust was present at all.
The commutator shows not even traces of erosion.

So--if I can cool it sufficiently, the new motor may be abusable to a fine and fast effect.

Any first thoughts? Some sort of positive displacement blower perhaps? Aquarium air pump?


Thanks for helping me kluge.
E-biking is an awful lot of fun for this newbie to that hobby.


Thanks,
Reid

This message has been edited. Last edited by: Reid Welch,

From the Unite Motor webpages

Unite_Motor.htm (44 Kb, 9 downloads)

It's a chain drive to the rear hub.
This is based on the Currie patent "electrodrive" system.

.

I agree with your conclusions that the failure is most likely due to overheating and you most likely could get more capacity by providing increased cooling. Some small improvement might be gained by trying to improve the radiative/convective characteristics of the existing motor surface (finned heat sinks for example) but not much. The biggest bang for the buck would be your idea of a fan to send external air through the inside of the motor. I think it will work and give you a pretty good increased capacity as long as the windings brushes and commutator aren't badly affected by moisture and contaminants in the air... no easy way to know without trying.

You might try posting this question at eng-tips.com ... a lot of special-purpose motor guys are there.

Out of curiosity, do you strap three car batteries to your bike for this thing?

Info overload, sorry.

The present goal: to force cool this motor
so that it might survive indefinite periods of overload (high current, high speed).

--
future question: Regarding the nature of windings=I know nothing.

IF we wanted to redesign this motor, and could only alter the winding, what wire and winding count would make it -more nearly right- for running at higher rpm with the present 36V supply?

-would we wind fewer turns of heavier wire?
(is that the general way?)

-if so, I presume it'd move increase the (presently lousy) efficiency at the desired, new, speed and loading.

- I will be cooling this motor in any event.
So, first order is to jerry up a neat, cheap-enough method to blow air through the motor, right?

-Later on I could learn to rewind the burned armature, if I might learn here some general rules of that discipline.

Thanks again,
Reid

This message has been edited. Last edited by: Reid Welch,

This bike, the (Currie) IZIP 24V cruiser, cost under $400 delivered.
Here it is with the third 12AH SLA brick lashed to a seat post rack.


--it is a fun thing--
and it keeps the car at home.


I will make its motor last despite overloading it.

I don't think rewinding a motor is an easy thing to pick up. Even harder to think you will somehow do a better job to squeeze more copper into those slots or optimize the design to make better use of the same copper than the OEM did. (Sometimes it's done for large motors by using newer thinner insulation materials).

For $85 I would be inclined to buy a new motor. You have the 250W model. I see they also offer some higher watt models. Maybe one of those would work? My first step would be to try to find a manufacturer's rep or application guide to guide me through selection of the best motor.

But I understand also sometimes it's just more fun to do it yourself...

quote:

Originally posted by electricpete:
I agree with your conclusions that the failure is most likely due to overheating and you most likely could get more capacity by providing increased cooling. Some small improvement might be gained by trying to improve the radiative/convective characteristics of the existing motor surface (finned heat sinks for example) but not much. The biggest bang for the buck would be your idea of a fan to send external air through the inside of the motor. I think it will work and give you a pretty good increased capacity as long as the windings brushes and commutator aren't badly affected by moisture and contaminants in the air... no easy way to know without trying.

You might try posting this question at eng-tips.com ... a lot of special-purpose motor guys are there.

Out of curiosity, do you strap three car batteries to your bike for this thing?

HI electricpete (great screen name!)
I posted without having seen your thoughts.

Now by photo you see that the bike in stock form
carried two 12AH 12V gel cells inside the oblong box.

I added the third battery (on the seatpost rack) as you see, in series with the factory batteries.

I'd considered painting the housing flat black for better emissivity. But, the motor, being in the direct rush of air at 25mph, probably wouldn't cool much better even with thin black paint on instead of the heat-retaining frosted silver (it's clearcoated diecast Al.)

Yeah--good... forced air appeals to you too.

If cool windings work better (lower resistance),
hah: consider the whacky possibility of dumping 12oz cans of refrigerant though the motor by manually operated valve and tubing with a pinhole "expansion" valve into the motor. A lossy but c o o l way to cool! Oh, but brushes like to work with some H20 present in the air, so that's not so good an idea; only Rube Goldberg-ish.

---Let's see! Lots of great engineers are here.
I am full of untrained, untutored enthusiasm.

-I also like old motors.
I'll post a picture of one from my collection of junk, if it's OK to digress that much. But later on.

Thanks very much Pete,

Reid

quote:

Originally posted by electricpete:
I don't think rewinding a motor is an easy thing to pick up. Even harder to think you will somehow do a better job to squeeze more copper into those slots or optimize the design to make better use of the same copper than the OEM did. (Sometimes it's done for large motors by using newer thinner insulation materials).

For $85 I would be inclined to buy a new motor. You have the 250W model. I see they also offer some higher watt models. Maybe one of those would work? My first step would be to try to find a manufacturer's rep or application guide to guide me through selection of the best motor.

But I understand also sometimes it's just more fun to do it yourself...

Well, at this time this gearmotor is the only one suitable for this application; Unite makes many motors, but nothing else with integral gearcase, to bolt to an ebike. I think, suppose, they have made this under license of the Currie patent (though I do not understand how such a basic system could be patented here in the first place).

Thanks. As you already offered, it is more fun to explore the limits of what we have.
And to keep the form factors small; to avoid an even larger, heavier lump hanging off the side of the bike, is good.

If not for the heat factor I think this'd hold up just fine.

The most pleasant finding was to learn that the brushes and commutator seem to be up to the high current challenge.

Cheers,
Reid

quote:

Originally posted by Reid Welch:
General question about wire enamels of today.

What are they at basis?

From the ancient era of single and double silked magnet wire; the very early varnished wire;

Phenolic resin enamels (I think) came in around 1912 or so.

The phenolics were pretty good--but correct my rusty recall: does the aging/breakdown of wire enamel involve hydrolysis?

(I think that's the term)

moisture, electrical pressures, decomposition.

Nylon was used (is used today?) for some wire enamels. I presume (knowing nothing because I'm not trained) that nylon is a low temperature coating. It's also hygroscopic.

-Are enamels today not immune to moisture/voltage degradation?

-I see small motors only. No longer do the makers impregnate the windings with varnish.
Is that due to cost contraints?

-Is a vacuum impregnation always preferable?

-What are these varnishes of today based upon?

Pardon me if you can for just having dumbed down this thread. But it seems to be about finished otherwise.

Thanks,

Reid

The following quotes are taken from "Electrical Insulation for Rotating Machines" by Greg Stone, IEEE Press 2002.

"The most common magnet wire for random-wound
stators in use today is a round copper wire insulated with a polyamide–imide insulation (Class 220°C) or polyester with a polyamide–imide overcoat. The insulation thickness is usually from 0.05 mm to about 0.1 mm. The most common standard covering magnet wires is the U.S. NEMA MW1000. The insulation on the magnet wire serves as the coil turn insulation."

"Most random-wound stators are coated with a varnish or resin after the coils have been inserted in the slot. This coating imparts resistance to moisture and contamination (which lead to electrical tracking), and also improves the electrical breakdown strength of the windings. Since NEMA MW1000 specifications for magnet wire do allow a certain number of “pinholes” in the insulation per length of wire, the varnish or resin ensures that partly conductive films cannot lead to turn-to-turn faults. In addition, the varnish or resin will improve the transmission of heat from the copper to the stator core, since the number of air pockets is reduced. As IFDs permeate the market, the voltage surges they create can lead to destructive PD in any air pockets. Thus filling of the air pockets with varnish or epoxy is becoming more critical, since PD can only occur if air pockets exist."

This doesn't directly address moisture resistance, but mentions there will be some defects in the wire which present a vulnerability in presence of conducting film (moisture plus dirt), unless impregnated over the magnet wire.

Great, thanks for the varnish information, Pete.

That clinches it for me: I'll obtain a small quantity of baking varnish from some local motor shop, if there is one here in town.

I can vacuum impregnate the armature myself--low tech: an old refrigerator compressor and an old pressure cooker pot= easy stuff.

I guess vac.impregnation might be overkill for such a small thing but I like the idea of pulling out as much entrained water vapor as possible (that pinhole effect the data mentioned.)

I see the motor will suffer extreme vibrations and rapid thermal cycling. The mechanical locking of a good varnish job would help reliable-ize this small motor.

--TODAY: found a suitable blower (hooray for small victories)

So--I'm well on track now. Just need to obtain a copy of any one of these Nidec blowers.

Thanks electricpete!


Reid

Here's the huskiest of the Gamma 30 line

http://www.nidec.com/newfans/newfans.htm


Learned today:
Here's an input from an e-scooter fellow; has used Gamma 30 fans for his conversion kits

http://tinyurl.com/mo97d

(all this has come in since I started this thread here the other day).

I will be following Howard Penrose's basic advice--I'll be venting with the aim to jet the air onto the commutator. Exit holes for this air at the other end of the motor.

Will have some sort of basic filtering/bug and water baffling.

And from the previous page of my thread at VisforVoltage.com,
a fellow advised about his having done it this way.
http://tinyurl.com/qzeuj

Hair drier Style


"You need an adequate amount of air flow to optimize the cooling. A simple fan will not work in most cases. An aquarium pump does not deliver nearly enough volume. Think hair dryer. A hair dryer dissipates over 1000w and doesn't burn up (unless you block the air flow). I used a small centrifugal brushless blower from a surplus place. It draws about 0.2 amps. A centrifugal blower is able to supply a good amount of air through a restricted passage."
excerpt from pictures/text at
http://tinyurl.com/qzeuj

--
This'll be easy and fun
Reminds me of the 1887 C&C motor
What would Mr. Crocker have done?


(humor)

Update.

It was a five week delay before a replacement bike motor arrived.

This thread details the adentures of force cooling the new gearmotor.

http://tinyurl.com/ngul5

I've come to a preliminary conclusion which I'll state here for consideration:

Relative humidity greatly affects commutation quality.

See the test run reports just made, if curious.

The bike motor thrives on damp air.
I do not now believe sealed brushed motors to be a sound propostion at all. They all should be ventilated and humidified if and as needed.
I do not suppose brushes work best with just a 'trace' of moisture; rather a surfeit of moisture in air-over constancy.

This is an empirically derived hypothesis,
and it's newly conceived.

Please set me straight if I am wrong.

Cheers and thanks,
Reid

Discovery
probably not new. But it's new to me)

To recap the electric bike motor story:

-The first OEM gearmotor burned up because I run it 100% past design center.

-The replacement is identical. However, I have ventilated the sealed motor case,
and added a small centrifugal blower to push air through the unit. Pictures are installed at the linked page.

-Early trials suggest a good cooling effect has been obtained.
monitoring the case temp shows that on a damp night (over 80% relative humidity) the case temperature does not exceed 108F.

-However, today is a dry day. A similar hard run resulted in a case temperature of 135F,
and air-exit temperature of nearly 150F.
I can guess that the armature is considerbly hotter.

So, why does the motor run cool in humid weather, and hot in dry weather?,
-ambient temperature being the same in both instances-.

Is this a well known effect?

Tony (Aerowhatt) gives a plausible answer at our ebike forum:
http://tinyurl.com/ykxtgu


What do you think?

Thanks,
Reid

This message has been edited. Last edited by: Reid Welch,

I have heard the same thing for transformers which have radiators or coolers cooled by external fans. If you hold the ambient temperature the same and increase the humidity, the transformer oil temperature would go down.

I remember I was surprised when I heard that, but a competent mechanical engineer explained it to me, he was very confident of that behavior, and it made sense to me at the time. I don't remember all the details of that explanation anymore. One thing for sure, water has higher specific heat capacity, so if we could get the same lbm/hr of moist air as dry air, we would have more cooling. But I tend to think of a fan as a constant-cfm device when we vary density. So if we decrease the density (moister air), that would decrease the mass flow rate for a given volumetric flow-rate which acts in the opposite direction (provides less cooling). Maybe the first effect is a bigger effect than the second?

With all that said, my gut feel is that it's not as big an effect as you described. It'll be interesting to see if your temperature comes back down when it gets humid again. If it doesn't come down, then you have a degrading/dieing motor on your hands.

Reid,

What a great thread! I just found it.

You would be a real prize for any company that ran lots of electric motors.

You have a great passion and you have questions! Asking questions is a great way yo learn.

Super!

Cannot wait to read that you have either set the world electric bike distance record or speed record!

Did you ever re-wind that motor?

Terry O